Defektr

Defektr is a Bittensor subnet that produces a single commodity: production-grade, edge-deployable AI models for manufacturing visual quality control.

Miners compete to build the best defect detection models, scored on accuracy, speed, and robustness. Validators maintain curated benchmark datasets and objectively evaluate miner submissions through Yuma Consensus. Factory customers purchase top-ranked models and deploy them locally on edge hardware (Jetson, Coral TPU, industrial PCs) at production-line speed.

The result is a decentralized R&D engine for industrial computer vision, where global competition between miners continuously produces improving models, and market demand from real factories steers development toward what industry actually needs.

During the hackathon, our team designed a high-potential subnet within the Bittensor ecosystem focused on delivering a genuine “proof of intelligence” through a robust incentive and mechanism design. We developed a comprehensive subnet proposal outlining emission and reward logic, clear miner and validator roles, and safeguards against low-quality or adversarial behavior. The design included a high-level algorithm covering task assignment, submission, validation, scoring, and reward distribution. We defined miner responsibilities (input/output formats, performance metrics such as quality, speed, and accuracy) and validator evaluation methodology with aligned incentives and recurring scoring cadence. Beyond the technical architecture, we articulated the business logic, competitive landscape, and long-term market rationale, demonstrating why the use case is uniquely suited for a Bittensor subnet. Finally, we prepared all required deliverables: a formal subnet design document (PDF/Notion/GitHub), an explanation video, a public introduction post, and a business pitch deck outlining go-to-market strategy, early adoption incentives, and sustainable growth pathways.

Pre-seed